雾霾天气下, 航拍设备无法准确获取图像信息, 为解决此问题, 提出一种改进暗通道窗口与透射率修正的图像去雾算法。首先, 用超像素分割有雾图像得到景深一致的局部窗口, 在每个窗口内计算暗通道, 同时根据大气光特性结合超像素进行大气光估计; 然后, 通过引导滤波细化透射率, 并建立自适应容差机制来修正图像明亮区域的透射率; 最后, 反演大气散射模型还原清晰图像。实验结果证明,该算法所得结果图像细节清晰、颜色自然, 且能处理多类雾天图像, 鲁棒性更好, 与经典算法相比具有显著优势。

In a hazy environment,aerial photography equipment cannot accurately obtain image information.To solve this problem,an image dehazing algorithm with improved dark channel window and transmittance correction is proposed.Firstly,the hazy image is segmented by superpixel to obtain local windows with consistent depth of field,and the dark channel is calculated in each window.Meanwhile,the atmospheric light is estimated by using superpixel according to the characteristics of atmospheric light.Secondly,the transmittance is refined by guided filtering,and an adaptive tolerance mechanism is established to correct the transmittance of the bright area in the image.Finally,the atmospheric scattering model is inverted to restore a clear image.The experimental results show that the result of the algorithm is clear in detail and natural in color,and can handle multiple types of hazy images with better robustness.Compared with classical and novel algorithms,it has significant advantages.